Invocation and runtime¶
Call agents and inspect conversations, feedback, runs, and asynchronous resource operations.
Choose a task¶
Task |
Page |
Description |
|---|---|---|
Call the agent |
Get the Agent Card before invoking an agent. Then invoke through A2A or MCP. |
|
Agent sessions |
Conversation APIs manage agent conversation records and display metadata. |
|
Agent feedback |
Feedback APIs review saved agent run feedback and summary statistics. Feedback submission still occurs through agent runtime interactions. |
|
Runtime state and records |
Inspect saved agent runtime tasks, events, manifests, turn snapshots, and data slices. |
|
Agent resource operation |
Inspect or cancel resource operations still in progress. These records are separate from agent invocation runs. |
|
Agent catalog |
Choose an API for the resource you need to operate on, then prepare the identity, IDs, and parameters listed on its action page. |
|
Web applications |
Choose an API for the resource you need to operate on, then prepare the identity, IDs, and parameters listed on its action page. |
|
Artifacts |
Read, save, and maintain task artifacts. |
|
Agent home resources |
Read resources used by agent home guidance and examples. |
|
UI conversations |
Manage conversations, groups, and display states in the interface. |